void TWgtNet::PutRnd01Wgts() {
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
for (int e = 0; e < NI.GetOutDeg(); e++) {
NI.GetOutEDat(e) = TInt::Rnd.GetUniDev();
}
}
}
void TWgtNet::MulEdgeWgt(const double& MulBy) {
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
for (int e = 0; e < NI.GetOutDeg(); e++) {
NI.GetOutEDat(e) *= MulBy;
}
}
}
// actors collaboration graph
PUNGraph TImdbNet::GetActorGraph() const {
TIntPrSet EdgeSet;
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
if (NI().GetTy() == mtyActor) {
const int NId1 = NI.GetId();
for (int e = 0; e < NI.GetOutDeg(); e++) {
if (NI.GetOutNDat(e).GetTy() != mtyActor) {
TNodeI NI2 = GetNI(NI.GetOutNId(e));
for (int e2 = 0; e2 < NI2.GetInDeg(); e2++) {
if (NI2.GetInNDat(e2).GetTy() == mtyActor) {
const int NId2 = NI2.GetInNId(e2);
EdgeSet.AddKey(TIntPr(TMath::Mn(NId1, NId2), TMath::Mx(NId1, NId2)));
}
}
}
}
}
}
PUNGraph G = TUNGraph::New();
for (int i = 0; i < EdgeSet.Len(); i++) {
const int NId1 = EdgeSet[i].Val1;
const int NId2 = EdgeSet[i].Val2;
if (! G->IsNode(NId1)) { G->AddNode(NId1); }
if (! G->IsNode(NId2)) { G->AddNode(NId2); }
G->AddEdge(NId1, NId2);
}
return G;
}
/////////////////////////////////////////////////
// Weighted network
double TWgtNet::GetEdgeWgt() const {
double wgt = 0;
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
for (int e = 0; e < NI.GetOutDeg(); e++) {
wgt += NI.GetOutEDat(e);
}
}
return wgt;
}
void TWgtNet::PermOutEdgeWgt() {
TFltV WgtV;
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
WgtV.Clr(false);
for (int e = 0; e < NI.GetOutDeg(); e++) {
WgtV.Add(NI.GetOutEDat(e));
}
WgtV.Shuffle(TInt::Rnd);
for (int e = 0; e < NI.GetOutDeg(); e++) {
NI.GetOutEDat(e) = WgtV[e];
}
}
}
int TMultimodalGraphImplB::BFSTraversalOneHop(const TVec< TPair<TInt,TInt> >& StartingVertices) const {
int NumVerticesAndEdges = 0;
for (int i = 0; i < StartingVertices.Len(); i++) {
TNodeI NI = GetNI(StartingVertices.GetVal(i));
TIntV AdjacentModes = TIntV(); NI.GetAdjacentModes(AdjacentModes);
for (int ModeIdx = 0; ModeIdx < AdjacentModes.Len(); ModeIdx++) {
int ModeId = AdjacentModes.GetVal(ModeIdx);
for (int e = 0; e < NI.GetOutDeg(ModeId); e++) {
NumVerticesAndEdges += NI.GetOutNId(e, ModeId);
}
}
}
return NumVerticesAndEdges;
}
void TWgtNet::DelMinWgtNodes(const double MinWgt) {
printf("Deleting Min Wgt %g nodes\n", MinWgt);
printf(" (%d,%d) -->", GetNodes(), GetEdges());
TIntV DelNIdV;
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
double wgt = 0;
for (int e = 0; e < NI.GetOutDeg(); e++) {
wgt += NI.GetOutEDat(e);
}
if (wgt < MinWgt) { DelNIdV.Add(NI.GetId()); }
}
for (int d = 0; d < DelNIdV.Len(); d++) {
DelNode(DelNIdV[d]);
}
printf(" (%d,%d)\n", GetNodes(), GetEdges());
}
TMultimodalGraphImplB::TEdgeI TMultimodalGraphImplB::GetEI(const TPair<TInt,TInt>& SrcNId, const TPair<TInt,TInt>& DstNId) const {
TNodeI CurNode = TNodeI(NodeToModeMapping.GetI(SrcNId.GetVal2()), &Graphs, &NodeToModeMapping);
TIntV AdjacentModes = TIntV();
CurNode.GetAdjacentModes(AdjacentModes);
TPair<TInt,TInt> ModeIdsKey = GetModeIdsKey(SrcNId.GetVal1(), DstNId.GetVal1());
int CurAdjacentMode = 0;
for (int i=0; i<AdjacentModes.Len();i++) {
if (AdjacentModes.GetDat(i) == DstNId.GetVal1()) {
CurAdjacentMode=i;
break;
}
}
return TEdgeI(CurNode, EndNI(), Graphs.GetDat(ModeIdsKey).GetEI(SrcNId.GetVal2(), DstNId.GetVal2()).GetCurEdge(), AdjacentModes, CurAdjacentMode);
}
// set actor's language and country
void TImdbNet::SetActorCntryLangByMajority() {
// set language
TIntPrV NIdToVal;
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
if (! NI().IsActor()) { continue; }
IAssert(NI().GetLang() == 0); // no language set
IAssert(NI.GetInDeg() == 0); // actors point to movies
int Nbhs=0; TIntH LangCntH;
for (int e = 0; e < NI.GetOutDeg(); e++) {
LangCntH.AddDat(NI.GetOutNDat(e).GetLang()) += 1; Nbhs++; }
if (LangCntH.IsKey(0)) { Nbhs-=LangCntH.GetDat(0); LangCntH.GetDat(0)=0; }
LangCntH.SortByDat(false);
if (LangCntH.GetKey(0) == 0) { continue; }
if (LangCntH[0]*2 >= Nbhs) {
NIdToVal.Add(TIntPr(NI.GetId(), LangCntH.GetKey(0))); }
}
for (int i = 0; i < NIdToVal.Len(); i++) {
GetNDat(NIdToVal[i].Val1).Lang = NIdToVal[i].Val2; }
printf(" language set: %d\n", NIdToVal.Len());
int cnt=0;
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) { if (NI().IsActor() && NI().GetLang()==0) cnt++; }
printf(" Actors NO language: %d\n\n", cnt);
// set country
NIdToVal.Clr(true);
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
if (! NI().IsActor()) { continue; }
IAssert(NI().GetCntry() == 0); // no country set
IAssert(NI.GetInDeg() == 0); // actors point to movies
int Nbhs=0; TIntH CntryCntH;
for (int e = 0; e < NI.GetOutDeg(); e++) {
CntryCntH.AddDat(NI.GetOutNDat(e).GetCntry()) += 1; Nbhs++; }
if (CntryCntH.IsKey(0)) { Nbhs-=CntryCntH.GetDat(0); CntryCntH.GetDat(0)=0; }
CntryCntH.SortByDat(false);
if (CntryCntH.GetKey(0) == 0) { continue; }
if (CntryCntH[0]*2 >= Nbhs) {
NIdToVal.Add(TIntPr(NI.GetId(), CntryCntH.GetKey(0))); }
}
for (int i = 0; i < NIdToVal.Len(); i++) {
GetNDat(NIdToVal[i].Val1).Cntry = NIdToVal[i].Val2; }
printf(" country set: %d\n", NIdToVal.Len());
cnt=0;
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) { if (NI().IsActor() && NI().GetCntry()==0) cnt++; }
printf(" Actors NO country: %d\n\n", cnt);
}
TFlt TNEANetMP::GetWeightOutEdges(const TNodeI& NI, const TStr& attr) {
TNode Node = GetNode(NI.GetId());
TIntV OutEIdV = Node.OutEIdV;
TFlt total = 0;
int len = Node.OutEIdV.Len();
for (int i = 0; i < len; i++) {
total += GetFltAttrDatE(Node.OutEIdV[i], attr);
}
return total;
}
void TMultimodalGraphImplB::RandomWalk(TVec< TPair<TInt,TInt> > NodeIds, int WalkLength) {
int CurrentLocalNodeId = NodeToModeMapping.GetKey(NodeToModeMapping.GetRndKeyId(TInt::Rnd));
TPair<TInt,TInt> CurrentNodeId = TPair<TInt,TInt>(NodeToModeMapping.GetDat(CurrentLocalNodeId), CurrentLocalNodeId);
int NodeIdIdx = 0;
NodeIds.SetVal(NodeIdIdx++, CurrentNodeId);
while (NodeIds.Len() < WalkLength) {
TNodeI NI = GetNI(CurrentNodeId);
TIntV AdjacentModes = TIntV(); NI.GetAdjacentModes(AdjacentModes);
// Throw an appropriately biased coin here
int EdgeId = TInt::Rnd.GetUniDevInt(NI.GetOutDeg());
int i;
for (i = 0; i < AdjacentModes.Len(); i++) {
int ModeOutDeg = NI.GetOutDeg(AdjacentModes.GetDat(i));
if (EdgeId < ModeOutDeg) { break; }
EdgeId -= ModeOutDeg;
}
NodeIds.SetVal(NodeIdIdx++, TPair<TInt,TInt>(AdjacentModes.GetDat(i), NI.GetOutNId(EdgeId, AdjacentModes.GetDat(i))));
}
}
void TNEGraph::Dump(FILE *OutF) const {
const int NodePlaces = (int) ceil(log10((double) GetNodes()));
const int EdgePlaces = (int) ceil(log10((double) GetEdges()));
fprintf(OutF, "-------------------------------------------------\nDirected Node-Edge Graph: nodes: %d, edges: %d\n", GetNodes(), GetEdges());
for (TNodeI NodeI = BegNI(); NodeI < EndNI(); NodeI++) {
fprintf(OutF, " %*d]\n", NodePlaces, NodeI.GetId());
fprintf(OutF, " in[%d]", NodeI.GetInDeg());
for (int edge = 0; edge < NodeI.GetInDeg(); edge++) {
fprintf(OutF, " %*d", EdgePlaces, NodeI.GetInEId(edge)); }
fprintf(OutF, "\n out[%d]", NodeI.GetOutDeg());
for (int edge = 0; edge < NodeI.GetOutDeg(); edge++) {
fprintf(OutF, " %*d", EdgePlaces, NodeI.GetOutEId(edge)); }
fprintf(OutF, "\n");
}
for (TEdgeI EdgeI = BegEI(); EdgeI < EndEI(); EdgeI++) {
fprintf(OutF, " %*d] %*d -> %*d\n", EdgePlaces, EdgeI.GetId(), NodePlaces, EdgeI.GetSrcNId(), NodePlaces, EdgeI.GetDstNId());
}
fprintf(OutF, "\n");
}
void TWgtNet::PermEdgeWgt() {
TFltV WgtV;
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
for (int e = 0; e < NI.GetOutDeg(); e++) {
WgtV.Add(NI.GetOutEDat(e));
}
}
WgtV.Shuffle(TInt::Rnd);
int w = 0;
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
for (int e = 0; e < NI.GetOutDeg(); e++) {
NI.GetOutEDat(e) = WgtV[w++];
}
}
}
// Wgt == -1 : take the weight of the edge in the opposite direction
void TWgtNet::AddBiDirEdges(const double& Wgt) {
TIntPrV EdgeV;
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
for (int e = 0; e < NI.GetOutDeg(); e++) {
if (! IsEdge(NI.GetOutNId(e), NI.GetId())) {
EdgeV.Add(TIntPr(NI.GetOutNId(e), NI.GetId())); }
}
}
for (int e = 0; e < EdgeV.Len(); e++) {
if (Wgt != -1) {
AddEdge(EdgeV[e].Val1, EdgeV[e].Val2, Wgt);
} else { // edge weight in the opposite direction
AddEdge(EdgeV[e].Val1, EdgeV[e].Val2, GetEDat(EdgeV[e].Val2, EdgeV[e].Val1));
}
}
}
PMultimodalGraphImplC TMultimodalGraphImplB::ConvertToImplC() const {
PMultimodalGraphImplC G = TMultimodalGraphImplC::New();
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
G->AddNode(TPair<TInt,TInt>(NI.GetModeId(), NI.GetId()));
}
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
int OutDeg = NI.GetOutDeg();
int InDeg = NI.GetInDeg();
G->ReserveInAndOutNIdV(TPair<TInt,TInt>(NI.GetModeId(), NI.GetId()), OutDeg, InDeg);
TVec<TInt> AdjacentModes = TVec<TInt>(); NI.GetAdjacentModes(AdjacentModes);
for (int i = 0; i < AdjacentModes.Len(); i++) {
int AdjacentMode = AdjacentModes.GetVal(i);
TPair<TInt,TInt> ModeIdsKey = GetModeIdsKey(NI.GetModeId(), AdjacentMode);
G->AddOutNIds(TPair<TInt,TInt>(NI.GetModeId(),NI.GetId()), Graphs.GetDat(ModeIdsKey).GetOutNId(NI.GetId()), AdjacentMode);
G->AddInNIds(TPair<TInt,TInt>(NI.GetModeId(),NI.GetId()), Graphs.GetDat(ModeIdsKey).GetInNId(NI.GetId()), AdjacentMode);
}
}
return G;
}
// set Language and Country for movies that do not have the value set
// for every movie find the mojority language/country in 1-hop neighborhood and set it
void TImdbNet::SetLangCntryByMajority() {
// set language
while (true) {
TIntPrV NIdToVal;
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
if (NI().GetLang() != 0) { continue; }
int Nbhs=0; TIntH LangCntH;
for (int e = 0; e < NI.GetOutDeg(); e++) {
LangCntH.AddDat(NI.GetOutNDat(e).GetLang()) += 1; Nbhs++; }
for (int e = 0; e < NI.GetInDeg(); e++) {
LangCntH.AddDat(NI.GetInNDat(e).GetLang()) += 1; Nbhs++; }
if (LangCntH.IsKey(0)) { Nbhs-=LangCntH.GetDat(0); LangCntH.GetDat(0)=0; }
LangCntH.SortByDat(false);
if (LangCntH.GetKey(0) == 0) { continue; }
if (LangCntH[0]*2 >= Nbhs) {
NIdToVal.Add(TIntPr(NI.GetId(), LangCntH.GetKey(0))); }
}
if (NIdToVal.Empty()) { break; } // done
for (int i = 0; i < NIdToVal.Len(); i++) {
GetNDat(NIdToVal[i].Val1).Lang = NIdToVal[i].Val2; }
printf(" language set: %d\n", NIdToVal.Len());
}
int cnt=0;
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) { if (NI().GetLang()==0) cnt++; }
printf(" NO language: %d\n\n", cnt);
// set country
while (true) {
TIntPrV NIdToVal;
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
if (NI().GetCntry() != 0) { continue; }
int Nbhs=0; TIntH CntryCntH;
for (int e = 0; e < NI.GetOutDeg(); e++) {
CntryCntH.AddDat(NI.GetOutNDat(e).GetCntry()) += 1; Nbhs++; }
for (int e = 0; e < NI.GetInDeg(); e++) {
CntryCntH.AddDat(NI.GetInNDat(e).GetCntry()) += 1; Nbhs++; }
if (CntryCntH.IsKey(0)) { Nbhs-=CntryCntH.GetDat(0); CntryCntH.GetDat(0)=0; }
CntryCntH.SortByDat(false);
if (CntryCntH.GetKey(0) == 0) { continue; }
if (CntryCntH[0]*2 >= Nbhs) {
NIdToVal.Add(TIntPr(NI.GetId(), CntryCntH.GetKey(0))); }
}
if (NIdToVal.Empty()) { break; } // done
for (int i = 0; i < NIdToVal.Len(); i++) {
GetNDat(NIdToVal[i].Val1).Cntry = NIdToVal[i].Val2; }
printf(" country set: %d\n", NIdToVal.Len());
}
cnt=0;
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) { if (NI().GetCntry()==0) cnt++; }
printf(" NO country: %d\n\n", cnt);
}
// IN-OUT edges are swapped (so that the prog runs faster)
// Send message via IN edge proportional to the OUT edge weight
void TWgtNet::ReinforceEdges(const int& NIters) {
THash<TInt, TFlt> OutWgtSumH;
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
double wgt = 0;
for (int e = 0; e < NI.GetOutDeg(); e++) {
wgt += NI.GetOutEDat(e); }
OutWgtSumH.AddDat(NI.GetId(), wgt);
}
printf("Reinforcing edges for %d iterations\n", NIters);
// iterate
TExeTm ExeTm;
for (int iter = 0; iter < NIters; iter++) {
for (TNodeI NI = BegNI(); NI < EndNI(); NI++) {
const double X = TInt::Rnd.GetUniDev() * OutWgtSumH.GetDat(NI.GetId());
double x = 0; int e = 0;
for ( ; x + NI.GetOutEDat(e) < X; e++) {
x += NI.GetOutEDat(e); }
IAssert(IsEdge(NI.GetOutNId(e), NI.GetId()));
GetEDat(NI.GetOutNId(e), NI.GetId()) += 1; // reinforce the edge
OutWgtSumH.GetDat(NI.GetOutNId(e)) += 1;
}
if (iter % (NIters/100) == 0) {
printf("\r%d [%s]", iter, ExeTm.GetStr());
}
}
printf(" done.\n");
}
void TNEANet::Dump(FILE *OutF) const {
const int NodePlaces = (int) ceil(log10((double) GetNodes()));
const int EdgePlaces = (int) ceil(log10((double) GetEdges()));
fprintf(OutF, "-------------------------------------------------\nDirected Node-Edge Network: nodes: %d, edges: %d\n", GetNodes(), GetEdges());
for (TNodeI NodeI = BegNI(); NodeI < EndNI(); NodeI++) {
fprintf(OutF, " %*d]\n", NodePlaces, NodeI.GetId());
// load node attributes
TIntV IntAttrN;
IntAttrValueNI(NodeI.GetId(), IntAttrN);
fprintf(OutF, " nai[%d]", IntAttrN.Len());
for (int i = 0; i < IntAttrN.Len(); i++) {
fprintf(OutF, " %*i", NodePlaces, IntAttrN[i]()); }
TStrV StrAttrN;
StrAttrValueNI(NodeI.GetId(), StrAttrN);
fprintf(OutF, " nas[%d]", StrAttrN.Len());
for (int i = 0; i < StrAttrN.Len(); i++) {
fprintf(OutF, " %*s", NodePlaces, StrAttrN[i]()); }
TFltV FltAttrN;
FltAttrValueNI(NodeI.GetId(), FltAttrN);
fprintf(OutF, " naf[%d]", FltAttrN.Len());
for (int i = 0; i < FltAttrN.Len(); i++) {
fprintf(OutF, " %*f", NodePlaces, FltAttrN[i]()); }
fprintf(OutF, " in[%d]", NodeI.GetInDeg());
for (int edge = 0; edge < NodeI.GetInDeg(); edge++) {
fprintf(OutF, " %*d", EdgePlaces, NodeI.GetInEId(edge)); }
fprintf(OutF, "\n out[%d]", NodeI.GetOutDeg());
for (int edge = 0; edge < NodeI.GetOutDeg(); edge++) {
fprintf(OutF, " %*d", EdgePlaces, NodeI.GetOutEId(edge)); }
fprintf(OutF, "\n");
}
for (TEdgeI EdgeI = BegEI(); EdgeI < EndEI(); EdgeI++) {
fprintf(OutF, " %*d] %*d -> %*d\n", EdgePlaces, EdgeI.GetId(), NodePlaces, EdgeI.GetSrcNId(), NodePlaces, EdgeI.GetDstNId());
// load edge attributes
TIntV IntAttrE;
IntAttrValueEI(EdgeI.GetId(), IntAttrE);
fprintf(OutF, " eai[%d]", IntAttrE.Len());
for (int i = 0; i < IntAttrE.Len(); i++) {
fprintf(OutF, " %*i", EdgePlaces, IntAttrE[i]()); }
TStrV StrAttrE;
StrAttrValueEI(EdgeI.GetId(), StrAttrE);
fprintf(OutF, " eas[%d]", StrAttrE.Len());
for (int i = 0; i < StrAttrE.Len(); i++) {
fprintf(OutF, " %*s", EdgePlaces, StrAttrE[i]()); }
TFltV FltAttrE;
FltAttrValueEI(EdgeI.GetId(), FltAttrE);
fprintf(OutF, " eaf[%d]", FltAttrE.Len());
for (int i = 0; i < FltAttrE.Len(); i++) {
fprintf(OutF, " %*f", EdgePlaces, FltAttrE[i]()); }
}
fprintf(OutF, "\n");
}